Synthetic DNA poses biosecurity risks that curated databases and screening algorithms can minimize. Here’s how they work now and how they could improve.
Conservatively, it would require more than an order of magnitude in improved solid phase synthesis efficiency to reach 5000bp.
Moreover, it is completely irrelevant. It is much more practical to stitch smaller oligos synthesized by extant solid phase synthesizers together by traditional gene synthesis methods. Polio virus was made this way like 25 years ago. For this reason, I don't think there is any point in making an oligosynthesizer with antivirus software. Improving solid phase synthesis of oligos is mostly a fluidics problem and the tech seems to moves agonizingly slowly. In practice, most oligosynthesis companies I know of use archaic instruments anyways.
The real reason basically no one has used this technology to synthesize artificial viruses for terroristic purposes is that biological weapons are very expensive and not particularly effective at doing more than catching headlines. Synthetic biological weapons doubly so.
That said, I really enjoyed the writing and thought that went into this piece, especially the stuff about how screening algorithms work.
What do you think about the Evonetix platform? Discussed in this interview with the CTO, for example: https://www.evonetix.com/news/interview-matthew-hayes . They are doing chemical synthesis, as well as assembly and error correction, on a benchtop device (they advertise "gene-length DNA", but I'm not sure what their exact length limitations are). It sounds like they are innovating on the solid phase chemistry side of things, though!
In general, I agree that long custom synthesis is more likely to be developed via enzymatic synthesis (c.f. Ansa Biotechnologies) or automated assembly methods (c.f. BioXp devices, Ribbon Biolabs) than pure chemical synthesis, but I feel like progress is speeding up here!
I don't think the estimate that benchtop DNA synthesizers will be able to make 5,000 bp by chemical synthesis in 2-5 years is remotely accurate considering that the longest chemically synthesized oligo is only 1005 bases as of 2023 . https://www.genengnews.com/topics/genome-editing/worlds-longest-oligo-produced-using-de-novo-synthesis/
Conservatively, it would require more than an order of magnitude in improved solid phase synthesis efficiency to reach 5000bp.
Moreover, it is completely irrelevant. It is much more practical to stitch smaller oligos synthesized by extant solid phase synthesizers together by traditional gene synthesis methods. Polio virus was made this way like 25 years ago. For this reason, I don't think there is any point in making an oligosynthesizer with antivirus software. Improving solid phase synthesis of oligos is mostly a fluidics problem and the tech seems to moves agonizingly slowly. In practice, most oligosynthesis companies I know of use archaic instruments anyways.
The real reason basically no one has used this technology to synthesize artificial viruses for terroristic purposes is that biological weapons are very expensive and not particularly effective at doing more than catching headlines. Synthetic biological weapons doubly so.
That said, I really enjoyed the writing and thought that went into this piece, especially the stuff about how screening algorithms work.
What do you think about the Evonetix platform? Discussed in this interview with the CTO, for example: https://www.evonetix.com/news/interview-matthew-hayes . They are doing chemical synthesis, as well as assembly and error correction, on a benchtop device (they advertise "gene-length DNA", but I'm not sure what their exact length limitations are). It sounds like they are innovating on the solid phase chemistry side of things, though!
In general, I agree that long custom synthesis is more likely to be developed via enzymatic synthesis (c.f. Ansa Biotechnologies) or automated assembly methods (c.f. BioXp devices, Ribbon Biolabs) than pure chemical synthesis, but I feel like progress is speeding up here!